TradeHouse Fidelity Lab (overfit validator) ยท Buyer resources
What is backtesting in machine learning?
A 26-method integrity harness that checks whether a backtest is overfit, with a $99 one-time audit or $399 five-audit pack and monthly Starter $49 / Pro $149 tiers. It is aimed at traders who want an independent check before risking capital.
Backtesting evaluates a model on historical periods as if predictions had been made at the time. Fit preprocessing and model parameters only on earlier observations, then score later outcomes. Include data availability delays, retraining rules and transaction costs when the output drives trades.
What to verify before buying
- Untouched out-of-sample periods
- Total number of tried strategies
- Leakage controls and execution costs
- Reproducible validation evidence
Published product pricing
Free: $0 per month; Starter: $49 per month; Pro: $149 per month; One-time audit: $99 per audit; Five-audit pack: $399 per pack